A kind of virtual interaction system and its method of work based on inertia capturing technology
Technical field
The present invention relates to a kind of virtual interaction system and its method of work based on inertia capturing technology.
Background technology
Man-machine interface in general sense is as two entities being separated from each other, and interface operator and computer
As the carrier of information transmission, among operational order can only be input to computer by user, then computer is again corresponding letter
Breath is fed back with action.In such a exchange method, user can only contact with computer, and be difficult to carry out with the object studied
Spontaneously, communicate with each other on one's own initiative.
Unmanned plane inspection at present has turned into an important means of polling transmission line.During inspection, unmanned plane carries
Camera is flown, and by flying flight attitude of the hand on ground by remote control control unmanned plane, is worked by ground control station
Personnel control camera angle in unmanned plane by computer and are taken pictures and be stored in the storage card of aircraft.Terminate in flight
The image in storage card is reached into backstage afterwards, carries out the judgement of defect by the identification to image from the background.The work is used at present
Pattern defect recognition rate is about 70% or so.Distance has man-machine inspection still to have larger gap.By analysis, cause discrimination low
The main reason for be:
1st, camera angle, focal length are limited by aircraft flight attitude, more numerous by ground control stand control camera
It is trivial, it can not often obtain optimal shooting angle;
2nd, in view of line security the reason for, unmanned plane need to keep at least 30 meters of safe distance with circuit, increase bat
The difficulty taken the photograph, it have impact on the definition of photo;
3rd, influenceed by factors such as shooting angle, distance, weather, image definition and resolution ratio can not fully meet automatic knowledge
Do not require, directly affects defect recognition rate.
And have man-machine inspection then be using helicopter carry patrol officer, maked an inspection tour by way of human eye+telescope circuit,
Identify defect.In such a mode, the tour free degree of people is bigger, and it is accurate that head steering is flexibly, positioning facilitates, and this is all nothing
It can not accomplish during man-machine inspection.So if it is flexible that camera shooting can be lifted while unmanned plane during flying mobility is kept
Property, it will the inspection efficiency and practicality of unmanned plane is substantially improved.
The content of the invention
The present invention is in order to solve the above problems, it is proposed that a kind of virtual interaction system and its work based on inertia capturing technology
Make method, the present invention is combined using virtual interacting technology and motion capture technology, can determined using the dexterous actions of human eye
Some module in potential field scape, it is interactive so as to decide whether to carry out with unmanned plane, while realize to the accurate of electric power line pole tower
Focusing and intelligent collection, realize effective support to the first visual angle inspection work.
To better illustrate technical scheme, first stated as follows, virtual interaction system is user and surrounding environment work
On the whole treated for one, form a virtual interactive environment, user can pass through sense organ, such as vision and tactile, with ring
Object within border carries out the interactive action of active.
To achieve these goals, the present invention adopts the following technical scheme that:
A kind of virtual interaction system based on inertia capturing technology, including virtual reality device, action model storehouse, image biography
Defeated module, image panorama concatenation module, head real-time control module and human-computer interaction module, wherein:
The virtual reality device, speed, position and the direction of action of moving target are configured to determine that, to catch manipulation
The inertia action of person, meanwhile, receive and show the image of human-computer interaction module transmission;
The action model storehouse, it is configured as being stored with the above-below direction action of head/gondola, the left and right of head/gondola
Direction action, the far and near direction action of head/gondola, the spinning movement of head/gondola, head/gondola location action and
The positioning of head/gondola or the template model of closing motion;
Described image transport module, it is configured as receiving and transmits the transmission line of electricity image of unmanned plane collection;
Described image panoramic mosaic module, it is configured as carrying out feature point extraction, Feature Points Matching to transmission line of electricity image
And image co-registration, obtain transmission line status panoramic picture;
The head real-time control module, it is configured as carrying out shaft tower, head and pod the correspondence between coordinate system
Relation and Parameter Switch, according to the panoramic picture of passback, image object offset is converted into inertia rotation data, passes through conversion
The image shift amount collected is converted into the Mechanical course amount of head and the controlled quentity controlled variable of gondola by parameter;
The human-computer interaction module, it is configured as gathering systematic parameter, real-time video and the defect location of unmanned machine head
Information, and these information are subjected to packing transmission according to corresponding form, issue the determination of head real-time control module head or/
With gondola control strategy scheme, intelligentized control method is carried out to it, and feed back information to virtual reality device.
Further, the virtual reality device includes inertia capture module, display module and sensor device, the biography
Sensor equipment, including but not limited to gyroscope, accelerometer and/or magnetometer, the current motion state of measurement moving target and
Angular speed state;The inertia capture module passes through its speed by being recorded to the motion of moving target in three dimensions
The degree physical message related to location parameter acquisition, and then the simulation of progress movement locus, the display module display image or
Video information.
Further, described motion state include forward, back, up, downwards, to the left, it is therein a kind of to the right or
Person is a variety of, and described angular speed state includes accelerating or slowed down.
Further, the virtual reality device also includes data processing unit, receives inertia capture module and sensor
The kinematics information that equipment collects, when target is in motion, change the positional information of moving target, so as to obtain moving target
Track, utilize inertial navigation principle complete moving target motion capture.
Further, the data processing unit carries out the transmission of real time data with action model storehouse, reads all shiftings
Dynamic action model, shift action model is combined with the action data captured, the matching with model is realized, so as to drive head
Motion, be finally model and catch Data Matching, and the action data one that model can and then be caught changes.
Further, described image transport module is compressed to view data changes with form, is ensureing that image is effective
Property with frame per second on the premise of, maximized compressed data communication port accounting, at ground receiver end, completed using the mechanism that decodes high
The recovery of clear image, in view data transmitting procedure, different control signals is transmitted using different data transmission links.
The virtual reality device is that VR or AR heads show equipment.
Method of work based on said system, comprises the following steps:
(1) kinematics information of moving target is collected using virtual reality device, completes the motion capture of moving target;
(2) spatial attitude of the moving target of determination and action are matched with the action model stored, identifies action
The action command that target is sent, and be sent to unmanned plane and control it to complete corresponding act;
(3) unmanned plane performs corresponding action, it is observed that visual line of sight in gather scene image, image is entered
Row takes out frame, quantization and decoding process;
(4) synthesis of high pressure shaft tower image is carried out using image mosaic technology, obtains ultrahigh resolution panoramic picture;
(5) Coordinate Conversion is carried out to head or/and gondola, according to the panoramic picture of passback, image object offset turned
Change inertia rotation data into, the image shift amount collected is converted into the Mechanical course amount and gondola of head by conversion parameter
Controlled quentity controlled variable, with realize keep unmanned plane during flying mobility while lifted IMAQ flexibility and real-time.
Further, in the step (1), the initial data that will collect, including acceleration magnitude and magnitude of angular velocity, utilize
Data anastomosing algorithm calculates object space attitude angle, including roll angle, the angle of pitch and/or course angle.
Further, in the step (3), image procossing specifically includes:On the premise of visual effect is not influenceed, to regarding
Frequency carries out taking out frame processing;The analog signal of image is quantified and encoded with 256 pixel values, completes data-signal processing,
And then to compensation deals in the processing of image progress inverse quantization, infra-frame prediction and frame, recover video data.
Further, in the step (4), feature point extraction is carried out to image, reads ultrahigh resolution high pressure shaft tower figure
Picture, and sampling diminution is carried out to image;Ultrahigh resolution image to be spliced is subjected to sampling diminution using bilinear interpolation, it is right
All imagery exploitation ORB algorithms after sampling is reduced carry out feature extraction.
Further, in the step (4), closest matching is carried out using the ORB features of extraction, passes through RASANC algorithms
Obtained matching double points are screened, obtain thick matching double points;Using the thick matching double points coordinate of extraction, calculate original
Respective coordinates in ultrahigh resolution image, and in the image block where the matching double points of original ultrahigh resolution image again
ORB features are extracted, are accurately matched.
Further, in the step (4), using being fade-in gradually to go out method and merge ultrahigh resolution adjacent image, obtain
To ultrahigh resolution panoramic picture.
Further, in the step (5), including:
(1) Coordinate Conversion:Shaft tower, unmanned aerial vehicle platform, pod are in three coordinate systems, and shaft tower and unmanned plane are put down
Platform is combined and changed, unmanned aerial vehicle platform and gondola are combined conversion, so as to three's directly progressive association conversion;
(2) amount of spin is determined:It is accurate to find transmission of electricity according to the tour video information of the existing real-time display of virtual reality device
Circuit parts and defective locations, navigate to the position of part and defect, and start virtual unit focusing function, obtain in real time virtual
The Inertia information of equipment, it is mapped to by Coordinate Conversion in gondola coordinate system, obtains the Mechanical course amount of gondola;
(3) communication code of gondola Mechanical course amount is carried out based on unmanned aerial vehicle station encoder, controlled quentity controlled variable is by angle, position
Offset information is put, horizontal, the vertical and angular velocity information of gondola is converted into, controls link to be uploaded to gondola by data, drive
Dynamic gondola and the visible ray collecting device of carrying complete IMAQ.
Further, unmanned aerial vehicle (UAV) control platform configuration works as follows:
(1) open video path and set parameter and video window size, then initiating communications parameter, obtains nobody
Machine real-time video resource is simultaneously shown.
(2) by observing the dynamic of real-time video, according to detection pattern and user experience, judge whether to receive current position
Put;
(3) technical factor, external factor and the accidentalia of defect location during unmanned plane inspection are analyzed, identification is each
Operation flow node may influence the factor of defect location accuracy, formulate countermeasure;
(4) by the thinking habit of user, judge the position of head adjustment and unmanned plane is adjusted by video angle
The position of gondola;
(5) according to user action to the control of head travel direction and function control, meanwhile, it is also real-time to the gondola of unmanned plane
Control, realize the accurate focusing to electric power line pole tower and intelligent collection.
Compared with prior art, beneficial effects of the present invention are:
(1) present invention improves operating efficiency, cost-effective expenditure, improve operating efficiency in terms of robot inspection and
Work quality, cost of human resources is saved, huge economic benefit is brought for detection work.
(2) accelerate capacity for technological innovation, lift application level, capture the key skill of a collection of virtual reality interactive operation
Art, perceive, transmission, processing, using etc. technical field obtain important research achievement.
(3) realize that operation is intelligent, it is ensured that fortune inspection safety, promote user to use the smart machines such as robot, unmanned plane
Wish, promote the healthy and sustainable development of China's Robot industry, so as to reduce security incident, reduce robot error rate and make
Contribution.
(4) optimize Consumer's Experience, promote new technology spread application, user can pass through sense organ (vision and tactile) and environment
Within object carry out active interactive action, this interactive mode based on virtual reality, relative in traditional sense
Human-computer interaction interface technically make great progress.
Brief description of the drawings
The Figure of description for forming the part of the application is used for providing further understanding of the present application, and the application's shows
Meaning property embodiment and its illustrate be used for explain the application, do not form the improper restriction to the application.
Fig. 1 virtual interacting architecture diagrams;
Fig. 2 inertia actions gather and Controlling model flow chart;
The real-time control structure figure of virtual interacting plateform systems of the Fig. 3 based on inertia action capturing technology;
Virtual interacting platform hardware architecture design figures of the Fig. 4 based on inertia action capturing technology;
Fig. 5 H.264 code patterns;
H.264, Fig. 6 decodes figure.
Embodiment:
The invention will be further described with embodiment below in conjunction with the accompanying drawings.
It is noted that described further below is all exemplary, it is intended to provides further instruction to the application.It is unless another
Indicate, all technologies used herein and scientific terminology are with usual with the application person of an ordinary skill in the technical field
The identical meanings of understanding.
It should be noted that term used herein above is merely to describe embodiment, and be not intended to restricted root
According to the illustrative embodiments of the application.As used herein, unless the context clearly indicates otherwise, otherwise singulative
It is also intended to include plural form, additionally, it should be understood that, when in this manual using term "comprising" and/or " bag
Include " when, it indicates existing characteristics, step, operation, device, component and/or combinations thereof.
In the present invention, term as " on ", " under ", "left", "right", "front", "rear", " vertical ", " level ", " side ",
The orientation or position relationship of instructions such as " bottoms " are based on orientation shown in the drawings or position relationship, only to facilitate describing this hair
Bright each part or component structure relation and the relative determined, not refer in particular to either component or element in the present invention, it is impossible to understand
For limitation of the present invention.
In the present invention, term such as " affixed ", " connected ", " connection " should be interpreted broadly, and expression can be fixedly connected,
Can also be integrally connected or be detachably connected;Can be joined directly together, can also be indirectly connected by intermediary.For
The related scientific research of this area or technical staff, the concrete meaning of above-mentioned term in the present invention can be determined as the case may be,
It is not considered as limiting the invention.
As background technology is introduced, unmanned plane inspection in the prior art be present by angle, focusing and line security
Deng the limitation of many factors, for its mode of operation defect recognition rate about 70% or so, distance has man-machine inspection still to have larger difference
Away from deficiency, in order to solve technical problem as above, present applicant proposes a kind of virtual interacting system based on inertia capturing technology
System and its method of work, the invention provides three-dimensional visible interaction true to nature, intelligent, having interaction, scalability
Operating environment;Study VR/AR heads and show equipment to the head operational control technology of high-definition camera, realize that user uses inertia action
Some module come in positioning scene, it is interactive so as to decide whether to carry out with it;What research virtual reality was combined with unmanned plane hangs
Cabin technique of real-time control, realize the accurate focusing to electric power line pole tower and intelligent collection;The high speed for studying high-definition image passes
Transferring technology, realize effective support to the first visual angle inspection work.
A kind of typical embodiment, as shown in figure 1, a kind of virtual interaction system based on inertia capturing technology is user
On the whole treated as one with surrounding environment, using hardware facility and the software formation one of correlation it is virtual interact ring
Border, user can pass through the interactive action that sense organ (vision and tactile) carries out active with the object within environment.It is characterized in that:
VR heads show device, motion capture module, action model storehouse, image transmission module, image panorama concatenation module, head control in real time
Module, human-computer interaction module.
The VR heads show device, are a kind of all round computer graph technology, multimedia technology, sensor technology, man-machine friendship
The computer neck that a variety of scientific and technical comprehensive developments such as mutual technology, network technology, stereo display technique and emulation technology are got up
The latest equipment in domain.
The inertia capture module, it is by being recorded to the motion of object in three dimensions, passing through its speed, position
The related physical message of parameter acquiring such as put, and then carry out the simulation of movement locus.Using this technology, target needs are tracked
Show in VR heads and integrated accelerometer is worn in equipment, the inertial sensor equipment such as gyroscope and magnetometer, this is the dynamic of a whole set of
Make seizure system, it is necessary to which multiple components cooperate, it is made up of inertia device and data processing unit.
The inertial sensor equipment such as gyroscope, accelerometer and magnetometer:For measuring apparatus current motion state and angle
Speed state, described motion state include forward, back, up, downwards, to the left, one or more therein to the right, institute
The angular speed state stated includes accelerating or slowed down.
Data processing unit, it is the kinematics information collected using inertia device, when target is in motion, these yuan of device
The positional information of part is changed, and so as to obtain the track of target motion, can complete to move by inertial navigation principle again afterwards
The motion capture of target.
Data processing unit also with action model storehouse carry out real time data transmission, all shift action models we deposit
Storage is in action module database, in order to allow the action data captured to drive cradle head control, it would be desirable to by model with catching
The action data grasped combines, and realizes the matching with model, is finally model and seizure data so as to drive the motion of head
Match somebody with somebody, and the action data that model can and then be caught moves up.
Action model storehouse includes the above-below direction action of head/gondola, head/the left and right directions action of gondola, head/and hung
The far and near direction action in cabin, the spinning movement of head/gondola, the positioning of the location action of head/gondola and head/gondola,
Closing motion.
Image transmission module, it is by the image document of communication system passback, real-time exhibition during unmanned plane inspection
Transmission line status.The usual geographical position of transmission line of electricity is remote, and public communication network coverage rate is poor, and unmanned plane inspection logarithm
It is higher according to the quality of transmission, requirement of real-time, data transfer task can not be completed by public communication network.To realize high quality
Inspection communication, meet the display demand of VR equipment, the special data communication network of inspection need to be set up, complete the reality of video data
When transmit.
To realize the real-time Transmission of high-definition image, view data is compressed and changed with form, ensureing that image is effective
Property with frame per second on the premise of, maximized compressed data communication port accounting.At ground receiver end, completed using the mechanism that decodes high
The recovery of clear image.In view data transmitting procedure, and data such as unmanned aerial vehicle (UAV) control signal, cradle head control signal, using not
Same data transmission link, that is, open up special high-definition image transmission link.When carrying out the display of VR equipment, need to only show airborne
Hold the data of downlink.
Image panorama concatenation module, can not once be panned using common digital imaging apparatus and ultrahigh resolution
Image.It can smoothly be solved the above problems using image mosaic technology, successfully realize the conjunction of ultrahigh resolution high pressure shaft tower image
Into.Panorama Mosaic technology relates generally to feature point extraction, Feature Points Matching and the aspect of image fusion technology three, wherein feature
The extraction effect of point directly affects later image splicing effect.
Head real-time control module, shaft tower, unmanned aerial vehicle platform, pod are in three coordinate systems, completed to nothing
The control of man-machine gondola, first, calculate corresponding relation and conversion parameter between three coordinate systems;Secondly, according to the image of passback,
Image object offset is converted into inertia rotation data by VR equipment;Finally, it is by conversion parameter that the image collected is inclined
Shifting amount is converted into the Mechanical course amount of head.Cradle head control amount carries out logic coding by earth station, and by data transmitting
Transmit to airborne end, controlled quentity controlled variable is decoded by airborne machine code, completes the control of gondola.
Human-computer interaction module, by the high-speed transfer communication channel of high-definition image gather unmanned machine head systematic parameter,
Real-time video and defect location information, and these information are subjected to packing transmission according to corresponding form.Man-machine interactive platform is simultaneously
Head/gondola control strategy scheme is issued by the passage, intelligentized control method is carried out to it.
As shown in figure 3, a kind of method of work of the virtual interaction system based on inertia capturing technology, comprises the following steps:
Step 1: inertia action is caught, the data message of inertial sensor equipment is gathered, being tracked target needs in VR heads
Integrated accelerometer is worn in aobvious equipment, the inertial sensor equipment such as gyroscope and magnetometer, this is a whole set of motion capture
, it is necessary to which multiple components cooperate, it is made up of system inertia device and data processing unit, and data processing unit utilizes used
Property the kinematics information that collects of device, when target is in motion, the positional information of these components is changed, so as to obtain mesh
The track of motion is marked, the motion capture of moving target can be completed by inertial navigation principle again afterwards.
The initial data collected such as acceleration magnitude, magnitude of angular velocity is transferred in the minds of microprocessor core, by data
Blending algorithm calculates object space attitude angle such as roll angle, the angle of pitch, course angle (Eulerian angles representation) etc..Again will be numerous
Data summarization at sensor node is transferred in action model storehouse and handled.
Step 2: as shown in Fig. 2 action storage and Model Matching, all shift action models we be stored in action
In module database, in order to allow the action data captured to drive cradle head control, it would be desirable to by model and capture dynamic
Make data combination, the matching with model is realized, so as to drive the motion of head.Robot is by the information transmission of hazardous environment to control
Person processed, effector make various actions according to information, and motion capture system gets off motion capture, are real-time transmitted to robot simultaneously
It is controlled to complete same action.
Step 3: the collection and transmission of image, by video acquisition device gather it is observed that visual line of sight in
Scene image.The usual geographical position of transmission line of electricity is remote, and public communication network coverage rate is poor, and unmanned plane inspection passes to data
Defeated quality, requirement of real-time are higher, can not complete data transfer task by public communication network.To realize patrolling for high quality
Inspection communication, meets the display demand of VR equipment, need to set up the special data communication network of inspection, complete the real-time biography of video data
It is defeated.
To realize high definition of the transmission line of electricity image in VR equipment, quick displaying, using H.246 Image Compression to height
Clear image is compressed coding and decoding with video, as shown in Figure 5, Figure 6, mainly includes:
1) frame processing is taken out:On the premise of visual effect is not influenceed, video is carried out to take out frame processing;
2) image quantization:The analog signal of image is quantified with 256 pixel values, encoded, is completed at data-signal
Reason.
3) video decodes:To compensation deals in the processing of image progress inverse quantization, infra-frame prediction and frame, recover video data.
Step 4: splicing and the real-time exhibition of image, can not once be panned using common digital imaging apparatus and
The image of ultrahigh resolution.It can smoothly be solved the above problems using image mosaic technology, successfully realize ultrahigh resolution high pressure
The synthesis of shaft tower image.
(1) feature point extraction
Ultrahigh resolution high pressure shaft tower image is read first, and sampling diminution is carried out to image;Utilize bilinear interpolation
Ultrahigh resolution image to be spliced is subjected to sampling diminution.Then all imagery exploitation ORB algorithms after being reduced to sampling are carried out
Feature extraction.ORB features employ Oriented FAST feature point detections operators and Rotated BRIEF Feature Descriptors.
ORB algorithms not only have the Detection results of SIFT feature, but also with sides such as rotation, scaling, brightness change consistency
The characteristic in face, it is most important that its time complexity has than SIFT greatly to be reduced so that ORB algorithms splice in high-definition image
And there is very big application prospect in terms of real time video image splicing.
(2) Feature Points Matching
Closest matching is carried out using the ORB features of extraction, obtained matching double points are sieved by RASANC algorithms
Choosing, obtains thick matching double points.Using the thick matching double points coordinate of extraction, the correspondence in original ultrahigh resolution image is calculated
Coordinate, and extract ORB features again in the image block where the matching double points of original ultrahigh resolution image, carry out accurate
Match somebody with somebody.Finally calculate the transformation matrix H between adjacent image.
(3) image co-registration
Using being fade-in gradually to go out method and merge ultrahigh resolution adjacent image, ultrahigh resolution panoramic picture is obtained, is spelled
Binding beam.
Step 5: head controls in real time, the real-time control to head/gondola mainly includes following steps:
(4) Coordinate Conversion.Shaft tower, unmanned aerial vehicle platform, pod are in three coordinate systems, and shaft tower (is sat greatly
Mark) and unmanned aerial vehicle platform is combined and changed, unmanned aerial vehicle platform and gondola are combined conversion, so as to which three is directly progressive
Association conversion.
(5) amount of spin is determined.Operator shows the tour video information of real-time display according to virtual reality device, accurately
It was found that transmission line part and defective locations.The position of part and defect is navigated to, and starts virtual unit focusing function, in real time
The Inertia information of virtual unit is obtained, is mapped to by Coordinate Conversion in gondola coordinate system, obtains the Mechanical course amount of gondola.
(6) information transfer.The communication code of gondola Mechanical course amount, controlled quentity controlled variable are carried out based on unmanned aerial vehicle station encoder
By angle, position offset information, the horizontal, vertical of gondola, angular velocity information are converted to, controls link to upload by data
To gondola, gondola and the visible ray collecting device of carrying is driven to complete IMAQ.
Step 6: the concrete meaning and its working method of human-computer interaction module are as shown in Figure 4:
1) video is shown:The platform starting stage, it is first turned on video path and sets parameter and video window size, so
Initiating communications parameter afterwards, obtain unmanned plane real-time video resource and show.
2) information receives:User is by observing the dynamic of real-time video, according to user conventional detection pattern and its experience,
Judge whether to receive current position.
3) user's decision-making:Technical factor, external factor, the accidentalia of defect location during analysis unmanned plane inspection
Deng identifying that each operation flow node may influence the factor of defect location accuracy, formulate countermeasure.
4) thinking controls:By the thinking habit of user, judge the position of head adjustment and adjusted by video angle
The position of whole unmanned plane gondola.
5) cradle head control:The control of head is determined according to the headwork of user, and headwork is including above and below, a left side
Right and fore-and-aft direction movement, also include the action of the phenotype such as rotation, fixed, so as to the control of head travel direction, work(
Can control.
6) equipment controls:The control of head simultaneously, is also controlled the gondola of unmanned plane, realized to electric power line pole tower in real time
Accurate focusing and intelligent gather.
The preferred embodiment of the application is the foregoing is only, is not limited to the application, for the skill of this area
For art personnel, the application can have various modifications and variations.It is all within spirit herein and principle, made any repair
Change, equivalent substitution, improvement etc., should be included within the protection domain of the application.
Although above-mentioned the embodiment of the present invention is described with reference to accompanying drawing, model not is protected to the present invention
The limitation enclosed, one of ordinary skill in the art should be understood that on the basis of technical scheme those skilled in the art are not
Need to pay various modifications or deformation that creative work can make still within protection scope of the present invention.